On mercredi 7 Juillet 2004 08:28, Dr. Holger Putz wrote:
> > The algorithm is in the literature, so why prevent
> > the implementation of it in software?
>
> It is true, the algorithm is in the literature; unfortunately, the exact
> form of the potential and especially its parameters have not been published
> yet. These parameters (which have been extracted from the CSD structures by
> Dr. Hofmann throughout several years of work and extensive computer
> calculations) are the actual valuable information, though...
This is really strange. Of course, I do not know all of the details about
the un-publishable data. I see only two possibilities which could technically
justify (although it would still be stupid IMHO) such a restriction :
(i) if the parameters are not just a few numbers derived from the database,
but are a significant part of the database (like a list of all interatomic
distances in the database)
(ii) if part of the study was directly paid by the CCDC
But if neither (i) or (ii) are true, i.e. there are just a few parameters
derived from a bought copy of the database, then this is a clear abuse of the
data. What next ? Will we have to pay royalties every time a new structure is
solved using some information mined from the database ???
I understand the will to protect copying raw data from the database, but NOT
derived information. This is the kind of abuse of intellectual property that
is damaging to innovation and algorithms improvement, disregarding the
commercial/open-source nature of the software. I'm not even sure it's legal
to pretend 'owning' such parameters, unless we are in case (i) or (ii).
VIncent
--
Vincent Favre-Nicolin
Université Joseph Fourier
http://v.favrenicolin.free.fr
ObjCryst & Fox : http://objcryst.sourceforge.net
